Hacker News Reader

Lack of connection pooling for database connections means serverless architecture will never be used for serious data heavy apps.

AWS already solving this with Serverless Aurora.

not necessarily into this serverless thing but that doesn't seem like a valid complaint, use a global connection pooler like pgbouncer

Does e.g. Amazon's offering have a way to do that? Or is there still a cold start time? Where do you run the pgbouncer that is always running?

in AWS Fargate, of course...

Try it.